Accepted Manuscript
The Fabrication of a Bifunctional Oxygen Electrodes without Carbon Components for
Alkaline Secondary Batteries
Stephen W.T. Price, Stephen J. Thompson, Xiaohong Li, Scott F. Gorman, Derek
Pletcher, Andrea E. Russell, Frank C. Walsh, Richard G.A. Wills
PII:
S0378-7753(14)00251-1
DOI:
10.1016/j.jpowsour.2014.02.058
Reference:
POWER 18721
To appear in:
Journal of Power Sources
Received Date: 3 December 2013
Revised Date:
5 February 2014
Accepted Date: 15 February 2014
Please cite this article as: S.W.T. Price, S.J. Thompson, X. Li, S.F. Gorman, D. Pletcher, A.E. Russell,
F.C. Walsh, R.G.A. Wills, The Fabrication of a Bifunctional Oxygen Electrodes without Carbon
Components for Alkaline Secondary Batteries, Journal of Power Sources (2014), doi: 10.1016/
j.jpowsour.2014.02.058.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
The Fabrication of a Bifunctional Oxygen Electrodes without Carbon
Components for Alkaline Secondary Batteries
RI
PT
Stephen W.T. Pricea, Stephen J. Thompsona, Xiaohong Lib, Scott F. Gormanb,
Derek Pletchera, Andrea E. Russella, Frank C. Walshb, Richard G.A. Willsb,
Energy Technology Group, University of Southampton, Southampton, SO17 1BJ, UK
M
AN
U
b
Chemistry, University of Southampton, Southampton, SO17 1BJ, UK
SC
a
The fabrication of a gas diffusion electrode (GDE) without carbon components is
described. It is therefore suitable for use as a bifunctional oxygen electrode in alkaline
secondary batteries. The electrode is fabricated in two stages (a) the formation of a PTFEbonded nickel powder layer on a nickel foam substrate and (b) the deposition of a NiCo2O4
TE
D
spinel electrocatalyst layer by dip coating in a nitrate solution and thermal decomposition.
The influence of modifications to the procedure on the performance of the GDEs in 8 M
NaOH at 333 K is described. The GDEs can support current densities up to 100 mA cm-2 with
state-of-the-art overpotentials for both oxygen evolution and oxygen reduction. Stable
EP
performance during > 50 successive, 1 hour oxygen reduction/evolution cycles at a current
AC
C
density of 50 mA cm-2 has been achieved.
Keywords: Carbon free gas diffusion electrode (GDE), bifunctional electrode, oxygen
evolution/reduction.
1
ACCEPTED MANUSCRIPT
1.
Introduction
A successful renewable energy economy will require energy storage to manage the
time differences between generation and customer demand. One solution is offered by flow
batteries [1-3] although none of the systems extensively studied offer ideal behaviour and
zinc/air battery, the electrode reactions are:
negative electrode
charge
Zn(OH)42- + 2e-
Zn + 4OH-
positive electrode
charge
4OH- - 4e-
O2 + 2H2O
battery
2Zn(OH)4
2-
charge
discharge
M
AN
U
discharge
SC
discharge
RI
PT
economics. Secondary metal/air batteries, particularly zinc/air [4-6], merit development. In a
2Zn + O2 + 4OH- + 2H2O
(1)
(2)
(3)
The battery has an open circuit potential of ~ 1.65 V. Clearly, one requirement is a
bifunctional oxygen electrode, ie. an electrode that supports both oxygen evolution and
TE
D
reduction with low overpotentials.
A recent communication [7] described a novel procedure for fabrication of a
bifunctional oxygen electrode for alkaline secondary metal air batteries. In this procedure, a
nickel metal powder/PTFE gas diffusion electrode (GDE) is preformed within a nickel foam
EP
prior to the deposition of a catalyst layer by dip coating and thermal treatment. This paper
now reports the influence of the numerous parameters in the fabrication procedure on the
AC
C
performance of these electrodes.
While the choice of electrocatalyst to minimise overpotentials is clearly important, it
needs to be recognised that a gas diffusion electrode suitable as a bifunctional oxygen
electrode must have a series of additional properties. It must have a structure that provides an
effective barrier to crossover of both gas and electrolyte while permitting a high flux of
oxygen to the catalyst centres during discharge and release of oxygen to the gas side during
charge. Also there must be low resistance, current pathway between catalyst centres and
external contacts that prevents IR drops and an uneven current distribution. The target is to
design electrodes that have the physical and mechanical properties to allow their scale-up and
implementation in flow cells with electrode areas up to 1 m2.
2
ACCEPTED MANUSCRIPT
The design of the GDEs was based on a number of conclusions from preliminary
experiments and the literature.
(a) Nickel cobalt spinel, NiCo2O4 was selected as the electrocatalyst [8-14]. Preliminary
experiments showed that it gave overpotentials for both oxygen evolution and
reduction that were at least comparable to other electrocatalysts including precious
RI
PT
metals. However, it had the advantage that it was prepared in a simple procedure and
this could be achieved at a relatively low temperature where other components of the
GDEs were stable. This is essential for application of the spinel catalyst after the
formation of the gas diffusion layer.
SC
(b) Carbon materials within the GDEs (powder and paper) were avoided since the
literature has concluded that carbons corrode under the forcing conditions of oxygen
M
AN
U
evolution [15-18] and this was also our experience.
(c) The polymer selected as binder in the GDEs was PTFE. Cation conducting polymers
were considered unsuitable because of the key role of hydroxide ion in the battery
chemistry and no anion conducting polymers with the appropriate properties have been
located.
TE
D
The medium chosen was 8 M NaOH at 333 K. Sodium hydroxide is substantially cheaper
than potassium hydroxide and allows a significant increase in the zincate concentration
(1.2 M cf 0.5 M). The use of the elevated temperature leads to large decreases in the
overpotentials for the electrode reactions as well as increasing the solubility of the sodium
Experimental
AC
C
2.
EP
zincate; it is also a typical steady state temperature for a large scale electrolysis cell system.
2.1 Chemicals
Nickel powder (2–10 µm particle size determined by SEM) was supplied by Huizhou
Wallyking Battery Ltd, China. Two sources of nickel foam were used - Goodfellow Metals
(thickness 1.9 mm, 20 pores/cm) and Changsha Lyrun New Material Co. Ltd (thickness 1.6
mm, 43 pores/cm). Nickel nitrate (Aldrich, 99.999%), cobalt(II) nitrate (Aldrich, ≥ 98%),
sodium hydroxide (Fisher, 97%), polytetrafluoroethylene (PTFE, Aldrich, 60 wt% dispersion
in H2O) were used as received.
An electrode prepared from nickel cobalt spinel, NiCo2O4, powder was used for
comparison. The powder was prepared by a thermal decomposition procedure.
3
ACCEPTED MANUSCRIPT
Ni(NO3)2.6H2O (14.54 g) and Co(NO3)2.6H2O (29.1 g) were dissolved in methanol (100 cm3) and
heated at 338 K in fume cupboard to evaporate solvent. The dried powder sample was placed
into the Carbolite furnace in air at 648 K for 20 hours. The resulting black powder was
characterized by SEM, TEM, EDAX, XRD, BET analysis and particle size analysis and a
further paper [19] will discuss the data in detail as well as comparing spinel powders from
RI
PT
several preparation procedures. Here, we note that samples from repetition of the preparation
led to materials with well-defined XRD patterns characteristic of a spinel structure, a ratio of
Ni:Co close to 1:2 (determined by both EDAX and elemental analysis), BET surface area of
2.2 Preparation of Gas Diffusion Electrodes
SC
~10 m2 g-1 and particles sized below 5 µm.
The procedure for making the bifunctional oxygen gas diffusion electrodes (GDEs) is
M
AN
U
illustrated with a particular example. The first stage led to a porous nickel powder/PTFE
layer on nickel foam. The nickel foam (a disc, diameter 12 mm) was ultrasonicated in
acetone for 20 minutes, acid etched in 1 M HCl at 353 K for ~ 1 hour and then washed with
water and ultrasonicated in water for 15 minutes. Nickel powder (150 mg) and 60 % PTFE
solution (75 mg) were mixed with isopropanol (0.5 cm3) and water (0.5 cm3). The paste was
then ultrasonicated for 20 minutes and homogenised for 4 minutes to form an ink before
TE
D
drying to a paste with a ratio of Ni:PTFE of 10:3. The Ni/PTFE paste (200 mg wet weight ~
120-150 mg dry weight) was spread uniformly over the Ni foam disc and pressed in a Specac
hydraulic press at 1.5 kN cm-2 and 298 K for 30 s. The second step was to form the catalyst
layer. The nickel powder/PTFE coated nickel foam was soaked in a solution containing 1 M
EP
Ni(NO3)2 and 2 M Co(NO3)2 in 50/50 isopropanol/water, dried at 298 K for ~ 60 s and then
heat treated at 648 K in air for 10 min to form the NiCo2O4 spinel. The dip, dry and heat
AC
C
cycle was repeated 3 times before the sample was calcined at 648 K for 3 hours. The
procedure always had the two stages but a number of parameters within it (eg. ratio of
Ni:PTFE, loading, dipping solution, thermal treatment) were varied as set out in the results
section. For comparison, some GDEs were prepared where NiCo2O4 spinel powder was used
instead of Ni powder; these were fabricated in a single stage.
Figure 1 reports SEM images of the NiCo2O4 coated Ni powder/PTFE GDEs. Figure
1(a) shows the uniform surface of the electrodes exposed to the electrolyte when mounted in
the cell; an SEM images of the gas side clearly show that foam structure is maintained.
Figure 1(b) shows a cross section SEM image; there is 50 µm dense layer on the electrolyte
side and the fill of the foam decreases away for this surface to give open foam. Figure 2
4
ACCEPTED MANUSCRIPT
shows high resolution SEM images of the surface of the NiCo2O4 coated Ni powder/PTFE
GDE along with the Ni/PTFE layer before deposition of the spinel catalyst. It can be seen
that before the coating with catalyst, the surface consists of small angular particles of nickel
while after coating, rather unstructured, ‘fluffy’ patches of the spinel catalyst are dispersed
over the surface.
RI
PT
2.3 Electrochemical Experiments
Electrochemical experiments were carried out in a water jacketed glass cell (volume 200
cm3), see figure 3, with a GDE, a platinum gauze counter electrode and a laboratory prepared
Hg/HgO reference electrode placed inside a compartment with a Luggin capillary. The GDE
SC
disc (diameter 12 mm) was mounted inside a PTFE holder, see figure 3(b) with the NiCo2O4
coated Ni powder/PTFE layer adjacent to the electrolyte. The area of the GDE exposed to
M
AN
U
the electrolyte was 0.5 cm2 and electrical contact was made with a circle of Ni wire around
the perimeter of the disc on the gas side. A Grant TC120 recirculator with 5 litre reservoir
maintained the electrolyte temperature at 333 K. O2 was passed to the rear of the GDE with a
feed rate of 200 cm3 min-1, controlled via a flow meter. The electrolyte was 8 M NaOH at 333
K. Current cycling was carried out under galvanostatic control at current densities in the
range 10 – 100 mA cm-2. All current densities are based on the geometric area of the
TE
D
electrode (0.5 cm2) exposed to the electrolyte and gas compartments. Electrochemical
measurements were carried using an Autolab potentiostat/galvanostat, PGSTAT128N.
2.4 Other Instrumentation
EP
SEM images and EDAX data were obtained on a JSM 6500F Scanning Electron
Microscope. Powder XRDs were recorded on an Agilent Supernova XRD Diffractometer
with a Mo kα source. A small sample of powder was prepared in a 0.3 mm diameter glass
AC
C
capillary. Theoretical NiCo2O4 pattern produced using Crystal Maker v8.7.2 and Crystal
Diffract v5.2. Surface areas were determined with a Micromeritics – Gemini BET Instrument
using nitrogen as the gas.
3
Results and Discussion
Initially, the performance of a NiCo2O4 coated Ni powder/PTFE GDE and a NiCo2O4
powder/PTFE GDE during both O2 reduction and O2 evolution were compared in 8 M NaOH
and at a temperature of 333 K. In these experiments, a cathodic current was passed for 30
minutes, then an anodic current for the same period (corresponding to discharging and
charging of a battery respectively) and the potential of the GDEs vs a Hg/HgO reference
5
ACCEPTED MANUSCRIPT
electrode were recorded. Figure 4 shows the responses for three current densities. It can be
seen that following a short initial period, the potentials are always constant. Also the
difference in potentials for O2 reduction and evolution increase slightly with current density
due to increases in the overpotentials and IR drops. Clearly, there are significant
overpotentials associated with both O2 reduction and evolution but the differences in potential
RI
PT
between the two reactions, 710 mV and 770 mV for the spinel powder and spinel coated Ni
electrodes respectively at 50 mA cm-2, compares favourably with all other catalysts tested and
this included a number containing precious metals. It can also be seen that the spinel powder
electrodes performed slightly better than the spinel coated Ni GDEs but the difference in
SC
performance decreased with increasing current density. The improvement in performance is,
however, achieved at the cost of a very heavy loading of spinel catalyst. Hence, electrodes
were prepared with mixtures of nickel metal and spinel powders, followed by dip coating and
The overpotentials for both O2 reduction and
M
AN
U
some results are summarised in table 1.
evolution decrease as the spinel content is increased but even 10% spinel corresponds to a
high weight loading of catalyst. Hence, since a significant improvement in performance
requires a high loading of spinel powder, the electrodes based on Ni powder have been
developed further.
TE
D
It was shown in an earlier communication [7] that O2 evolution and reduction occur
with the transition metals ions within the spinel structure in different oxidation states. This
leads to a different open circuit potential following reduction or evolution and to the initial
periods before the potential takes up a constant value when the electrode is switched between
EP
O2 evolution and reduction (see figure 4). The first reaction to occur is always the change in
oxidation state of the metals ions within the spinel before the O2 reactions take over when the
AC
C
spinel is full oxidised/reduced. As expected the initial period is most pronounced with higher
spinel content (powder vs coating) and lower current density.
Table 1 also reports the behaviour of a Ni powder/PTFE GDE that has not been
subjected to the second stage of the fabrication procedure, ie. dip coating and heat treatment
to form a spinel coating. The resulting GDE is a very bad electrode for O2 reduction. In the
steady state with a cathodic current, the potential takes up a value close to – 1000 mV vs
Hg/HgO with the major reaction occurring being H2 evolution. The NiCo2O4 layer is the
catalyst for the O2 reactions and a key component of the bifunctional GDEs.
6
ACCEPTED MANUSCRIPT
The conditions for the formation of the spinel coating in the GDEs were investigated.
In all cases, the thermal treatment was carried out at 648 K for 3 hours since NiCo2O4 is
formed rapidly at this temperature and thermogravimetric analysis showed that PTFE
underwent decomposition above 670 K. Indeed, the fact that the spinel is formed at a
relatively low temperature where the PTFE is chemically stable is a key factor in the choice
RI
PT
of the catalyst for these GDEs. The temperature of 648 K is, however, certainly sufficient to
cause the polymer to flow and the heat treatment is likely to be influential in determining the
final structure of the GDEs. Firstly, GDEs were prepared using several different nickel and
cobalt salts dissolved in water and the results are presented in table 2. The significant
SC
difference is in the potential for the reduction reaction since the nickel foam substrate is
already a reasonable catalyst for O2 evolution. The temperature of the thermal treatment is
insufficient to convert the chloride material to the spinel structure but both the acetates and
M
AN
U
the nitrates are converted to effective catalyst and the latter were used in all later preparations.
Secondly, the total concentration of the Ni/Co nitrate solution in the dip solution was varied
(0.75, 1.5 and 3 M) while maintaining the Ni:Co ratio at 1:2. The overpotentials for both O2
reduction and evolution were reduced using the more concentrated solution; GDEs prepared
with the 3 M solution gave a potential gap between O2 reduction and evolution ~ 100 mV
TE
D
less than GDEs prepared with the 0.75 M solution with a current density of 50 mA cm-2.
Finally, the influence of the solvent for the dip solution was studied. Solutions of the Ni/Co
nitrates were prepared in 1:1 mixtures of water with five alcohols. The alcohols led to small
improvements in the performance of the GDEs probably resulting from changes in surface
EP
tension and/or viscosity leading to differences in the ability of the dip solution to diffuse into
the Ni powder/PTFE layer. The 1:1 isopropanol:water dip solution led to a GDE with the
AC
C
smallest difference in potential between O2 reduction and evolution. With the 1:1
isopropanol:water solution containing 3 M Ni/Co nitrates, three dip/thermal treatment cycles
were sufficient to give the optimum GDE performance when the spinel catalyst loading was
estimated by weight change as ~ 3 mg cm-2.
The ratio of PTFE binder to nickel powder is expected to be a key factor in
determining the performance of the GDEs, particularly the stability of performance during
extended operation. Electrodes with PTFE contents between 12 % and 27 % were therefore
tested. All the GDEs initially supported both reactions. Figure 5 reports the separation of the
potentials for O2 evolution and reduction as a function of PTFE content; the data is taken
from the initial charge/charge cycles at 50 mA cm-2. At the extremes of this range, it was
7
ACCEPTED MANUSCRIPT
difficult to obtain reproducible data but over the range 18 – 23% electrodes could be
fabricated reproducibly and the separation of the potentials for O2 evolution and reduction
varied little. The problem at low PTFE content was entry of the electrolyte into the porous
structure and this became more obvious with cycling of the electrodes – the potential
difference between O2 evolution and reduction increased and on dismantling the cell,
RI
PT
electrolyte could be seen on the gas side. With PTFE content above 23 %, the potential
difference was higher either because of hindered access of gas or electrolyte to the catalyst
centres or an increase in the resistance of the electrodes. Hence, long term stability without a
large increase in the potential difference was considered critical, 23 % PTFE (ratio of Ni
SC
powder:PTFE = 10:3) was selected as the safe option.
Further sets of experiments were carried out to define the influence of the weight
M
AN
U
loading of Ni powder/PTFE (10:3) paste and the pressure exerted during the manufacture of
the GDEs. In the first set the loading was varied over the range 120 - 280 mg cm-2 of wet
paste. The lowest loading suffered from electrolyte leakage through to the gas side but over
the range 140 - 280 mg cm-2 good responses were obtained, see figure 6, with a small trend to
higher overpotentials with thickening of the Ni powder/PTFE layer. Electrodes were also
pressed with compressions in the range 0.5 – 2.5 kN cm-2; the lowest pressure was
TE
D
insufficient to create a compact electrodes and leakage was an immediate problem, but above
1 kN cm-2 the electrodes gave good data with a slight tendency for the overpotentials to
increase with increasing pressure (see figure 7), probably due to a decrease in size of the gas
pores. Electrodes prepared with a loading of 200 mg cm-2 wet paste (approximately 90 mg
EP
cm-2 Ni + 30 mg cm-2 PTFE) and a compression of 1.5 kN cm-2 showed reproducible
behaviour and good long term stability and have been used for further electrode development.
AC
C
Several electrodes were further tested by cycling 10 times between 30 minutes O2
reduction and 30 minutes O2 evolution using current densities of 20 mA cm-2 or 50 mA cm-2.
In general, stable potentials were observed each cycle following initial periods where the
spinel coating is oxidised/reduced and commonly the potentials for O2 reduction and O2
evolution were unchanged during the 10 cycles. Some electrodes, eg. those with a low PTFE
content, showed a small negative shift in the potential during O2 reduction over the 10 cycles
and this was taken as an indication of ingress of aqueous electrolyte into the GDE structure
reducing the access of O2 gas to the catalyst sites.
8
ACCEPTED MANUSCRIPT
In the literature, air is preferred to oxygen as the cell feed when batteries are
discharged. Although the possibility exists for storing and using the pure O2 evolved during
charge (and this avoids CO2 degradation of the alkaline electrolyte when non-purified air is
employed), some tests employed an air feed. The responses during charge/discharge cycling
were similar in shape and stable potentials were established but the potential for oxygen
RI
PT
reduction shifted negative. At 50 mA cm-2, the negative shift was 70 mV.
These carbon free GDEs also met the requirement that during charge the oxygen was
evolved into the gas phase and away from the interelectrode gap. While a continuous and
rapid stream of fine H2 gas bubbles can be seen leaving the Pt mesh cathode, no O2 gas
SC
bubbles appear at the GDE/electrolyte interface. Hence, while with the present equipment,
any small fraction of the O2 entering the electrolyte cannot be quantified, it is clear that
M
AN
U
almost all the O2 is released to the back of the GDE.
Although the glass cell and auxiliary equipment used in this programme were not
designed for testing over many days, some longer term experiments have been carried out.
Figure 8 shows the potential vs time responses for cycles 1 – 10 and 40 - 50 for an
experiment carried out with a current density of 50 mA cm-2 and fabricated using the precise
TE
D
conditions of the recipe in the Experimental Section. It can be seen that there is a slight trend
to an improvement in performance with cycling but the changes in the potentials over the 50
cycles are small. The average separation of the potentials for O2 reduction and evolution
during the 50 cycles is 800 mV. Such performance in a Zn/air battery is equivalent to a
EP
voltage efficiency ~ 60 %. In fact, the experiment was continued for 100 cycles but beyond
50 cycles, the potential for O2 reduction became more negative with time, reaching -440 mV
AC
C
after 100 cycles.
Larger GDEs have been fabricated. 100 mm x 100 mm electrodes made from the
thinner Ni foam have been tested in a flow cell. These larger electrodes are flat with a
uniform black appearance and their thickness was measured as 0.75 mm. They are also robust
and can be extensively handled without damage making them well suited to incorporation
into flow cells.
4.
Conclusions
Gas diffusion electrodes without carbon components have been satisfactorily
fabricated and tested as a bifunctional electrode for O2 reduction/evolution. The electrodes
9
ACCEPTED MANUSCRIPT
can operate at 100 mA cm-2 and cycle well at 50 mA cm-2 and the performance may well be
adequate for practical secondary Zn/air batteries. These current densities are, however, low
compared to those possible with MEA type GDEs with carbon components for fuel cells ( >
1000 mA cm-2) suggesting that further enhancement of the O2 flux is still possible. Moreover,
the overpotentials observed for both O2 reduction and O2 evolution, while comparable with
RI
PT
those reported for other catalysts, are not as low as one would like. The choice of NiCo2O4
spinel as the catalyst results from the fact that for this method of fabrication of the GDEs, the
catalyst layer must be prepared at a temperature below that where the PTFE decomposes.
Acknowledgement
SC
5.
Financial support by the European Commission (Theme 2010.7.3.1) Energy Storage
acknowledged.
6.
1.
References
256759, is gratefully
M
AN
U
Systems for Power Distribution Networks, Grant Agreement No.
P. Leung, X. Li, C. Ponce de Leon, L. Berlouis, C.T.J. Low and F.C. Walsh, Progress in
Redox Flow Batteries, Remaining Challenges and Their Applications in Energy Storage,
2.
TE
D
RSC Advances, (2012) 1-32.
Z. Yang, J. Zhang, M.C.W.Kintner-Meyer, X. Lu, J.P. Lemmon and J. Lui,
Electrochemical Energy Storage for Green Grid, Chem. Rev., 111 (2011) 3577-3613.
3.
M. Skyllas-Kazacos, M.H. Chakrabarti, S.A. Hajimolana, F.S. Mjalli and M. Saleem,
R55-R79.
R.P. Hamlen, T.B. Atwater, in: D. Linden, T.B. Reddy (Eds.), Handbook of Batteries,
AC
C
4.
EP
Progress in Flow Battery Research and Development, J. Electrochem. Soc., 158 (2011)
third ed., McGraw-Hill, New York, 2002, pp. 38.1-38.53.
5.
F. Cheng and J. Chen, Metal Air Batteries: From Oxygen Reduction Electrochemistry to
Cathode Catalysts, Chem. Soc. Rev., 41 (2012) 2172-2192.
6.
J. Pan, L. Ji, Y. Sun, P. Wan, J. Cheng, Y. Yang and M. Fan, Preliminary Study of
Alkaline Single Flowing Zn–O2 Battery, Electrochem. Commun., 11 (2009) 2191-2194.
7.
X. Li, D. Pletcher, A.E. Russell, F.C. Walsh, R.G.A. Wills, S.F. Gorman, S.W. Price and
S.J. Thompson, A Novel Bifunctional Oxygen GDE for Alkaline Secondary Batteries,
Electrochem. Commun., 34 (2013) 228.
10
ACCEPTED MANUSCRIPT
8.
K. Kinoshita, Electrochemical Oxygen Technology, John Wiley & Sons Inc. 1992.
9.
L. Jörissen, Bifunctional Oxygen/Air Electrodes, J. Power Sources, 155 (2006) 23-32.
10. M.R. Tarasevich, B.N. Efremov, in: S. Trasatti (Ed.), Electrodes of Conductive Metallic
Oxides, Elsevier Scientific Publishing Company, 1980, pp. 221-59.
RI
PT
11. P. Rasiyah, A.C.C. Tsueng, D.B. Hibbert, A Mechanistic Study of Oxygen Evolution on
NiCo2O4. 1. Formation of Higher Oxides, J. Electrochem. Soc., 129 (1982) 1724-1727.
12. P. Rasiyah and A.C.C. Tseung, A Mechanistic Study of Oxygen Evolution on NiCo2O4.
SC
2. Electrochemical Kinetics, J. Electrochem. Soc., 130 (1983) 2384-2386.
13. P.N. Ross, Proceedings of the 10th Annual Battery Conference on Applications and
M
AN
U
Advances, 1995, pp. 131.
14. X. Li, F.C. Walsh and D. Pletcher, A Comparison of Nickel Based Electrocatalysts for
Oxygen Evolution in a Zero Gap Water Electrolyser with a Hydroxide Conducting
Membrane, Phys. Chem. Chem. Phys. 13 (2011) 1162-1167.
15. N. Staud, and P.N. Ross, The Corrosion of Carbon Black Anodes in Alkaline Electrolyte,
TE
D
Part II. Acetylene Black and the Effect of Oxygen Evolution Catalysts on Corrosion, J.
Electrochem. Soc., 133 (1986) 1079-1084.
16. P.N. Ross and M. Sattler, The Corrosion of Carbon Black Anodes in Alkaline Electrolyte,
EP
Part III. The Effect of Graphitization on the Corrosion Resistance of Furnace Blacks, J.
AC
C
Electrochem. Soc., 135 (1988) 1464-1470.
17. N. Staud, H. Sokol and P.N. Ross, The Corrosion of Carbon Black Anodes in Alkaline
Electrolyte, Part IV. Current Efficiencies for Oxygen Evolution from Oxide-Impregnated
Graphitized Furnace Blacks, J. Electrochem. Soc., 136 (1989) 3570-3576.
18. S. Műller, F. Holzer, H. Arai and O. Haas, A Studyof Carbon-Catalyst Interaction in
Bifunctional Air Electrodes for Zinc-Air Batteries, J. New Mater. Electrochem. Systems,
2 (1999) 227-232.
19. S.W. Price, S.J. Thompson, D. Pletcher and A.E. Russell, to be published.
11
100 % Ni
100% Ni
90% Ni/10% NiCo2O4
50% Ni/50% NiCo2O4
100% NiCo2O4
Dip
Coated
No
Yes
Yes
Yes
Yes
Reduction E/ mV
vs Hg/HgO
H2 evolution
-208
-183
-166
-138
Evolution E/ mV
vs Hg/HgO
***
658
628
589
532
SC
Sample
RI
PT
ACCEPTED MANUSCRIPT
ORR-OER
Gap, ∆E/mV
866
811
755
670
M
AN
U
Table 1 Potentials for O2 reduction and evolution at 50 mA cm-2 for GDEs prepared with Ni
powder (dip coated with NiCo2O4 and also without dip coating), spinel and mixtures
of nickel and spinels (dip coated). GDEs were 10:3 Ni powder:PTFE and the spinel
Nitrate
Sulphate
Chloride
Acetate
O2 Reduction E/ mV
vs Hg/HgO
-125
-378
-1068
-149
EP
Salt
TE
D
coating was formed at 648 K for 3 hours. 8 M NaOH. 333 K.
O2 Evolution E/ mV
vs Hg/HgO
594
575
526
600
ORR-OER Gap,
∆E/mV
719
953
1594
749
AC
C
Table 2 Influence of the anion of the nickel and cobalt salts used in the dip coating solution
on performance. GDEs were 10:3 Ni powder:PTFE and the spinel coating was
formed at 648 K for 3 hours. Constant current experiments carried at 20 mA cm-2in
8 M NaOH at 333 K.
12
ACCEPTED MANUSCRIPT
Legends to Figures
Figure 1
SEM images of the bifunctional GDE (a) the surface of the spinel coated Ni powder/PTFE
layer and (b) a cross section of the electrode.
RI
PT
Figure 2
High resolution SEM images of the surfaces of the bifunctional GDE (a) Ni powder/PTFE
layer before coating with catalyst (b) NiCo2O4 coated Ni powder/PTFE layer.
SC
Figure 3
Schematic of the three electrode cell used together with an expansion of the GDE structure.
M
AN
U
Figure 4
Comparison of the performance of a NiCo2O4 coated Ni powder/PTFE GDE and a NiCo2O4
powder/PTFE GDE during O2 reduction and O2 evolution. Potential vs time plots at (a) 20
mA cm-2 (b) 50 mA cm-2 and (c) 100 mA cm-2. 8 M NaOH. 333 K.
Figure 5
TE
D
The influence of the PTFE content on the performance of NiCo2O4 coated Ni powder/PTFE
GDEs for O2 reduction and O2 evolution at 50 mA cm-2 - difference in potentials for O2
reduction and O2 evolution vs % PTFE. 8 M NaOH. 333 K.
Figure 6
AC
C
Figure 7
EP
Plots of potential versus time for NiCo2O4 coated Ni powder/PTFE GDEs during O2
reduction and O2 evolution at 20 mA cm-2as a function of the loading with Ni powder/PTFE
(10:3) paste. 8 M NaOH. 333 K.
Influence of compression during NiCo2O4 coated Ni powder/PTFE GDE preparation on O2
reduction, O2 evolution cycling at 50 mA cm-2. 8 M NaOH, 333K, 200 ml min-1 O2,
Figure 8
Potential vs time responses during current density cycling of a NiCo2O4 coated Ni
powder/PTFE GDE at 50 mA cm-2 in 8 M NaOH at 333 K. Oxygen feed rate: 200 cm3 min-1.
Shown are 1-10 and 40-50 cycles.
13
M
AN
U
SC
RI
PT
ACCEPTED MANUSCRIPT
(a)
(b)
Figure 1 SEM images of the bifunctional GDE (a) the surface of the spinel coated Ni
AC
C
EP
TE
D
powder/PTFE layer and (b) a cross section of the electrode.
Figure 2 High resolution SEM images of the surfaces of the bifunctional GDE (a) Ni
powder/PTFE layer before coating with catalyst (b) NiCo2O4 coated Ni
powder/PTFE layer.
14
ACCEPTED MANUSCRIPT
M
AN
U
SC
RI
PT
(a)
AC
C
EP
TE
D
(b)
Figure 3 Schematic of the three electrode cell used together with an expansion of the GDE
structure.
15
ACCEPTED MANUSCRIPT
0.8
(a)
0.57 V
0.6
O2 evolution
0.2
O2 reduction
0.0
-0.08 V
-0.10 V
-0.2
Spinel powder/PTFE GDE
Spinel coated Ni/PTFE GDE
-0.4
0.0
0.2
0.4
0.6
0.8
1.0
Time / h
SC
0.8
RI
PT
E vs Hg/HgO / V
0.52 V
0.4
(b)
0.61 V
0.6
0.2
M
AN
U
E vs Hg/HgO / V
0.57 V
O 2 evolution
0.4
O2 reduction
0.0
-0.14 V
-0.2
-0.16 V
Spinel powder/PTFE GDE
Spinel coated Ni/PTFE GDE
-0.4
0.2
0.4
0.6
TE
D
0.0
0.8
1.0
Time / h
0.8
0.66 V
(c)
0.62 V
EP
O 2 evolution
0.4
0.2
0.0
O 2 reduction
-0.2
-0.23V
AC
C
E vs Hg/HgO / V
0.6
-0.26 V
-0.4
Spinel powder/PTFE GDE
Spinel coated Ni/PTFE GDE
-0.6
0.0
0.2
0.4
0.6
0.8
1.0
Time / h
Figure 4 Comparison of the performance of a NiCo2O4 coated Ni powder/PTFE GDE and
a NiCo2O4 powder/PTFE GDE during O2 reduction and O2 evolution. Potential vs
time plots at (a) 20 mA cm-2 (b) 50 mA cm-2 and (c) 100 mA cm-2. 8 M NaOH. 333 K.
16
ACCEPTED MANUSCRIPT
960
940
920
900
880
860
840
820
800
10
12
14
16
18
20
22
24
26
28
SC
wt.% PTFE
RI
PT
-2
ORR-OER gap @ 50 mA cm / mV
980
M
AN
U
Figure 5 The influence of the PTFE content on the performance of NiCo2O4 coated Ni
powder/PTFE GDEs for O2 reduction and O2 evolution at 50 mA cm-2 - difference in
potentials for O2 reduction and O2 evolution vs % PTFE. 8 M NaOH. 333 K.
0.7
0.6
0.5
TE
D
E vs Hg/HgO / V
0.4
0.3
0.2
0.1
0.0
-0.1
-2
170 mg cm
-2
210 mg cm
-2
250 mg cm
-2
290 mg cm
EP
-0.2
-0.3
500
1000
1500
2000
2500
3000
Time / s
AC
C
0
Figure 6 Plots of potential versus time for NiCo2O4 coated Ni powder/PTFE GDEs during
O2 reduction and O2 evolution at 20 mA cm-2as a function of the loading with Ni
powder/PTFE (10:3) paste. 8 M NaOH. 333 K.
17
ACCEPTED MANUSCRIPT
0.7
0.6
0.4
1.0 (kN cm-2)
1.5 (kN cm-2)
2.0 (kN cm-2)
2.5 (kN cm-2)
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
0
500
1000
1500
2000
2500
3000
Influence of compression during NiCo2O4 coated Ni powder/PTFE GDE preparation
on O2 reduction, O2 evolution cycling at 50 mA cm-2. 8 M NaOH, 333K, 200 ml min-1 O2,
AC
C
EP
TE
D
M
AN
U
Figure 7
SC
Time / s
RI
PT
E vs Hg/HgO / V
0.5
Figure 8 Potential vs time responses during current density cycling of a NiCo2O4 coated
Ni powder/PTFE GDE at 50 mA cm-2 in 8 M NaOH at 333 K. Oxygen feed rate:
200 cm3 min-1. Shown are 1-10 and 40-50 cycles.
18
ACCEPTED MANUSCRIPT
Highlights:
RI
PT
The Fabrication of a Bifunctional Oxygen GDE without Carbon Components for Alkaline Secondary
Batteries
Fabrication and development of a carbon free bifunctional gas diffusion electrode
•
Good stability and overpotentials on cycling at 50 mA cm-2 for >50 cycles
•
Operational up to current densities of 100 mA cm-2
AC
C
EP
TE
D
M
AN
U
SC
•